Review Article Mind Body Interventions for Chronic Pain in Older Adults: A Structured Review

Transcription

1 Blackwell Publishing IncMalden, USAPMEPain Medicine American Academy of Pain Medicine? Original ArticlesMind Body Pain Older AdultMorone and Greco PAIN MEDICINE Volume 8 Number PAIN & AGING SECTION Review Article Mind Body Interventions for Chronic Pain in Older Adults: A Structured Review Natalia E. Morone, MD, MSc,* and Carol M. Greco, PhD *Department of Medicine, Division of General Internal Medicine and Department of Psychiatry, University of Pittsburgh, Pittsburgh, Pennsylvania, USA ABSTRACT ABSTRACT Study Design. We conducted a structured review of eight mind body interventions for older adults with chronic nonmalignant pain. Objectives. To evaluate the feasibility, safety, and evidence for pain reduction in older adults with chronic nonmalignant pain in the following mind body therapies: biofeedback, progressive muscle relaxation, meditation, guided imagery, hypnosis, tai chi, qi gong, and yoga. Methods. Relevant studies in the MEDLINE, PsycINFO, AMED, and CINAHL databases were located. A manual search of references from retrieved articles was also conducted. Of 381 articles retrieved through search strategies, 20 trials that included older adults with chronic pain were reviewed. Results. Fourteen articles included participants aged 50 years and above, while only two of these focused specifically on persons aged 65 years. An additional six articles included persons aged 50 years. Fourteen articles were controlled trials. There is some support for the efficacy of progressive muscle relaxation plus guided imagery for osteoarthritis pain. There is limited support for meditation and tai chi for improving function or coping in older adults with low back pain or osteoarthritis. In an uncontrolled biofeedback trial that stratified by age group, both older and younger adults had significant reductions in pain following the intervention. Several studies included older adults, but did not analyze benefits by age. Tai chi, yoga, hypnosis, and progressive muscle relaxation were significantly associated with pain reduction in these studies. Conclusion. The eight mind body interventions reviewed are feasible in an older population. They are likely safe, but many of the therapies included modifications tailored for older adults. There is not yet sufficient evidence to conclude that these eight mind body interventions reduce chronic nonmalignant pain in older adults. Further research should focus on larger, clinical trials of mind body interventions to answer this question. Key Words. Mind Body and Relaxation Techniques; Mind Body Relations (Metaphysics); Complementary Therapies; Aged; Pain Disorder Reprint requests to: Natalia E. Morone, MD, MSc, 230 McKee Place, Suite 600, Pittsburgh, PA 15213, USA. Tel: ; Fax: ; moronene@upmc.edu. Introduction C hronic nonmalignant pain is a common condition among the older population, occurring in up to 50% of community-dwelling older American Academy of Pain Medicine /07/$15.00/ doi: /j x

2 360 Morone and Greco adults [1]. It is associated with a significant amount of morbidity, such as depression [2,3], disability [4,5], and overall decreased quality of life [6]. Despite its widespread prevalence, many older adults live with pain because of inadequate treatment [7]. Reasons for this include lack of physician training in pain management in the older adult and increased susceptibility to medication side effects precluding their use [8,9]. Given the above, there is a compelling place for complementary mind body therapies for pain relief in older adults, because these therapies do not involve pharmacotherapy and can address the psychological and emotional burden of chronic pain. Most mind body therapies have been studied in younger populations. Although it is reasonable to assume that older adults would respond in the same way as younger people to mind body interventions, there is little evidence to support this. Additionally, we do not know whether mind body therapies would be inherently less (or more) acceptable to an older population, or whether they would have uniquely beneficial effects in older adults. The latter may be the case, because mind body therapies may have effects on function and cognition that may be independent from the effects on pain. This is particularly germane to the older adult, as maintaining functional independence is a high priority for them. Therefore, we considered it important to examine mind body therapies in more detail and conducted a review of mind body interventions for chronic nonmalignant pain in older adults. We were interested in the feasibility, safety, retention rates, and effect sizes for pain measures (when available) for the modalities reviewed. We chose the following eight mind body modalities to review: biofeedback, progressive muscle relaxation (PMR), meditation, guided imagery, hypnosis, tai chi, qi gong, and yoga. We chose these therapies because they are commonly used, have been studied in clinical trials, and are widely accepted mind body therapies. The current review does not include cognitive-behavioral therapy (CBT), which is typically considered a category of psychotherapy, or patient educational programs such as the Arthritis Self Management Program [10], due to their general acceptance within the biomedical system of care. Although these well-accepted programs contain components such as relaxation or guided imagery, the mind body components are not taught as extensively or as in-depth as they would be individually. The National Center for Complementary and Alternative Medicine defines mind body medicine in the following way: Mind-body medicine focuses on the interactions among the brain, mind, body, and behavior, and the powerful ways in which emotional, mental, social, spiritual, and behavioral factors can directly affect health. It regards as fundamental an approach that respects and enhances each person s capacity for self-knowledge and self-care, and it emphasizes techniques that are grounded in this approach. [11] The eight treatments we reviewed share the characteristics described in the National Institutes of Health (NIH) definition of mind body medicine. Each treatment emphasizes the importance of the patient s active participation in his or her health and well-being. Thus, these therapies require the patient to be self-motivated and selfdisciplined with a willingness to explore the emotional, social, spiritual, and behavioral context of his or her illness. The first step in this self-exploration usually begins by cultivating a calm and peaceful mental and physical state. This is usually engendered by slowing breathing and focusing attention on the breath, or by visualizing a tranquil place. If the therapy is a movement-based mind body therapy, this is usually done by synchronizing movement with breathing, or being mindful of body sensation during movement. These therapies take practice to become accustomed to them, and frequently home practice is suggested. They do not provide a quick-fix as might occur with taking a pain medication. They are also characterized by increased proficiency with time. Immediate effects of a session may include a decreased respiratory rate and heart rate as the relaxation response occurs. Otherwise, the mechanisms of action for mind body therapies are largely unknown. Group sessions are common among mind body therapies, and may include discussion that explores patients experience with the mind body technique and any insight discovered into the emotional, mental, social, spiritual, or behavioral dimensions of their condition. Methods We searched the MEDLINE (1966 March 2006), PsycINFO (1967 March 2006), AMED (1985 March 2006), and CINAHL (1982 March 2006) databases. Search terms used were: mind body and relaxation techniques, biofeedback, progressive muscle relaxation, meditation, mindfulness meditation, mindfulness, transcendental meditation, guided imagery, hypnosis, tai chi, tai ji, chi gong, yoga, aged, pain,

3 Mind Body Pain Older Adult 361 persistent pain, pain intractable, chronic disease, and older adults or older. We also manually reviewed the reference lists of retrieved articles. We included articles of randomized controlled clinical trials or uncontrolled clinical trials of older adults with chronic nonmalignant pain published in the English language. Articles were excluded if they: (1) did not study chronic nonmalignant pain; (2) were not published in English; (3) were not intervention trials; (4) were review or theoretical articles (e.g., not a primary study); and (5) did not include any older adults. We had intended to include only those studies with adults 65 years of age and older. However, this was not possible in many of the modalities studied, and so studies that included a younger population were also reviewed. We did not review the gray literature, such as abstracts, or other unpublished materials, such as dissertations. Studies were rated on quality (Table 1) based upon their consistency with CONSORT guidelines [12] and given a grade of high, moderate, low, or very low based on the guidelines formulated by the American College of Chest Physicians Task Force, which in turn are based on the guidelines formulated by the international GRADE group [13]. The grade is based on the confidence in the estimate of effect. For example, although a case series comparing older and younger persons response to interventions is valuable for assessing safety and feasibility, it would receive a very low grade for contribution to quality of evidence for the intervention, because it would not meet most CONSORT guidelines and any estimate of effect is very uncertain. Study quality is also reduced when randomization procedures, inclusion and exclusion criteria, and blinding of evaluators is not clearly described. When further research is very unlikely to change our confidence in the estimate of effect, it would receive a high grade. A moderate grade reflects that future research may change the estimate, and a low grade reflects that further research is very likely to change the estimate. Effect sizes comparing treatment and control group change scores were calculated where possible and are presented in Table 1. Results The database search strategies yielded a total of 381 articles on the following modalities: biofeedback and relaxation (73), transcendental meditation (TM) (70), mindfulness meditation (150), hypnosis (29), guided imagery (10), tai chi (4), qi gong (5), and yoga (40). Of the 381 articles retrieved through search strategies, 20 were deemed appropriate for review based upon the above criteria. Table 1 summarizes the 20 reviewed mind body intervention trials that included older adults with chronic pain. Fourteen articles included participants aged 50 years and above, while only two of these studies focused specifically on persons aged 65 years. An additional six articles included both persons under and over 50 years of age. Thirteen articles were controlled trials, of which 12 described random assignment to intervention. An additional five uncontrolled pre post trials were reviewed. The remaining two studies consisted of a clinical record review and a multiple baseline comparison of subject pairs. There is considerable overlap among the interventions described, and many are used in combination with one another. For example, PMR, imagery, and biofeedback are frequently combined. Of the identified studies in this structured review, seven clearly combined more than one modality, such as relaxation + imagery, or biofeedback + other relaxation techniques (see column 5 of Table 1 for intervention details). For clarity of presentation in the following study descriptions, we have categorized studies into relaxation-based, concentration-based, and movement-based mind body therapies. Relaxation-Based Mind Body Therapies: PMR and Biofeedback Training Biofeedback and other relaxation-based therapies are recognized as beneficial for chronic pain [14]. Although several reviews and randomized controlled trials [15 18] of biofeedback and various other relaxation-based therapies for various chronic pain conditions support the utility of these methods for reducing chronic low back pain, headache, and rheumatologic pain, few studies have specifically evaluated biofeedback and other relaxation methods for treating pain in the older adult. Progressive Muscle Relaxation Progressive muscles relaxation (PMR) is a systematic relaxation method developed in the late 1920s by Jacobson [19]. In PMR, patients tense and then let go the tension in various muscle groups, from one end of the body to the other. With practice over days or weeks, enhanced awareness of the body and increased ability to relax quickly and easily can be gained. Typically, patients are taught

4 362 Morone and Greco Table 1 Characteristics of included studies Study Design/ Methods 65 Years Condition/Age/N Relaxation Lundgren and Stenstrom 1999 [25] RCT No RA age 28 70, mean = 58 N = 68 Rybarczyk et al [27] RCT No Chronic illness (at least 1/3 with chronic pain) age 50 87, mean = 64.5 N = 178 Rybarzyk et al [26] RCT No Chronic illness (at least 1/3 with chronic pain) age 60 N = 243 Arena et al [24] Pre post No Headache age N = 10 Biofeedback Nicholson and Blanchard 1993 [29] Multiple baseline across subject pairs No Headache N = 14 Middaugh and Pawlick 2002 [35] Clinical record review No Chronic pain older younger N = 117 Intervention/Control/ Length of Treatment Outcome Measures PMR + imagery vs usual care 10 sessions AIMS 2 mobility, arm function* AIMS 2 pain Mind body wellness classes vs home video wellness program vs usual care: PMR, breathing, autogenics, imagery, + education 8 sessions Wellness classes vs wait-list control (see Rybarczyk et al. [25]) PMR 7 sessions MPQ-SF* Sleep* BAI* CES-D* *For class and home video vs usual care MSCL* for classroom vs usual care HPLP LSI MPQ-SF* (posttreatment but not 1-year follow-up) MSCL-Sleep* (postand 1 year) BAI* at post CES-D* at post 7/10 >50% reduction in headache at 3-month follow-up, 2 headache free PMR, EMG BF, CT, thermal BF for migraine patients/ symptom monitoring wait-list control 12 sessions PMR, EMG BF, cognitive therapy, PT 8 12 sessions 7/14 of BF and 2/7 of control >50% improvement in headache GDS* STAI-state* STAI-trait Pain VAS* (older patients pain reduction > younger patients) Pain Effect Sizes/ Quality Grade Notes Unable to compute effect size Quality: low Effect size MPQ-SF: Tx vs home video control 0.04 Tx vs wait-list 0.31 Quality: low 11% attrition in PMR 13% in control Attendance good but home practice low No group differences found at 1-year follow-up (15% attrition overall, 32% among classroom group) Effect size MPQ-SF Pre post 0.2 Pre-1 year 0.15 Quality: low 19% attrition: classroom, 20% attrition: control Younger persons more likely than older to drop out of treatment group Unable to compute effect size Quality: very low Attrition not reported Efforts were made to ensure that elderly subjects understood instructions Unable to compute effect size Quality: very low 26% attrition Treatment was adapted for elderly by increasing length of sessions and reducing muscle contraction level Unable to compute effect size Quality: very low Attrition not reported Nonstandardized multidisciplinary treatment Age effects on obtainable physiologic endpoints noted Program tolerated well by older patients

5 Mind Body Pain Older Adult 363 Study Kabela et al [31] Arena et al [30] Meditation Morone et al [43] Kabat-Zinn 1982 [42] Guided imagery Baird and Sands 2004 [47] Lewandowski 2004 [48] Hypnosis Gay et al [55] Design/ Methods 65 Years Condition/Age/N Pre post No Headache age N = 18 Pre post No Headache age N = 8 RCT Yes Chronic low back pain age 65 N = 37 Pre post No Chronic pain age N = 58 RCT Yes OA women age 65 N = 28 RCT No Chronic pain age 34 90, median = 61 N = 42 RCT No OA patients mean age 65 N = 36 Intervention/Control/ Length of Treatment Outcome Measures Pain Effect Sizes/ Quality Grade Notes PMR, some patients received BF and/or CT 8 12 sessions EMG BF 12 sessions 64% of sample had >50% improvement in headache 50% had >50% improvement in headache medication index 1-month follow-up 3 months following treatment, 4/8 >50% headache reduction, 2/8 had 35 45% headache reduction Effect size Headache Index: pre post 0.66 Quality: very low Effect size Headache Index: pre post 1.04 Quality: very low 11% attrition prior to treatment Unstandardized behavioral treatment, follow-up included Treatment tailored to increase comprehension Attrition not reported Treatment modified to ensure comprehension (e.g., longer sessions) Follow-up included MBSR program vs wait-list control 8 sessions MBSR 10 sessions CPAQ* SF-36 physical function* Pain Rating Index* Body Parts Problem Assessment MSCL* SCL-90-R GSI* Effect size CPAQ: pre post 0.83 Effect size: SF-36 Physical Function Scale: pre post 0.46 Quality: moderate Effect size Pain Rating Index: pre post 0.70 Quality: very low 32% attrition in intervention group, 6% attrition from control Included 3-month follow-up 12% attrition Guided imagery + PMR vs usual care 12 sessions Seven-minute guided imagery recording vs wait-list control 3 times/day over 4 days AIMS 2 pain* Self-reported mobility* Unable to compute effect size Quality: very low Pain VAS* Effect size Pain VAS: 1.14 Quality: low 4% attrition Adherence reported 3% attrition Short-term study (4 days), good treatment adherence noted PMR vs hypnosis vs usual care 8 sessions Pain VAS* Pain medication* (for hypnosis and relaxation groups compared with usual care) Effect size Pain VAS: 1.34 hypnosis vs control 0.59 hypnosis vs PMR 0.76 relaxation vs control Quality: moderate 12% attrition Pain reduction most rapid in hypnosis group, study included 6-month follow-up No instructions for home practice Pain medications decreased in hypnosis and relaxation groups

6 364 Morone and Greco Table 1 Continued Study Design/ Methods 65 Years Condition/Age/N Intervention/Control/ Length of Treatment Outcome Measures Pain Effect Sizes/ Quality Grade Notes Tai chi Hartman et al [57] Song et al [58] Kirsteins et al [82] RCT No Lower extremity OA average age 68 (range 49 81) N = 33 RCT No OA women age 55 N = 72 2 controlled studies, random assign. unclear No RA age N = 47 (study 1) N = 28 (study 2) Tai chi vs meeting and telephone control 24 hours of instruction Tai chi vs telephonecontact control 12 weeks Tai chi vs usual care Study 1: 10 sessions over 10 weeks Study 2: 10 sessions over 5 week. Satisfaction with health* Arthritis self-efficacy* Pain and functional measures** Joint pain and stiffness, physical function (K- WOMAC)* Joint tenderness, swollen joint count 50-foot walk* (study 2 only) Grip strength Effect size AIMS Pain: 0.79 Quality: moderate Effect size Joint Pain change scores 0.67 Quality: moderate Effect size Joint Pain: Study 1: 0.40 Study 2: % attrition One of few studies to include an attention control 43% attrition in tai chi, 39% attrition in control 50% attrition of controls, 20% attrition of tai chi of study 1 10% of usual care and 17% of tai chi attrition in study 2 Samples of the 2 studies were not independent Yoga Kolasinski 2005 [74] Sherman et al [77] Garfinkel et al [76] Garfinkel et al [80] Pre post No Knee OA mean age 58 (50 68) N = 11 RCT No Chronic low back pain mean age 44 N = 101 RCT No OA of hands age N = 25 RCT No Carpal tunnel syndrome age 24 77, median = 52 N = 51 Yoga 8 sessions Yoga vs exercise vs self-care book 12 sessions Yoga + relaxation + education vs wait-list 8 sessions Yoga vs usual care + wrist splint 8 sessions Pain and physical function (WOMAC)* RDS* Symptom bothersomness* Medication use* Pain (joint tenderness and hand pain during activity)* Pain VAS* Grip strength* Unable to compute effect size Quality: very low Unable to compute effect size Quality: high Effect size Pain with Dolorimeter: 1.57 right hand, 1.16 left hand Quality: low Effect size Pain VAS: 0.41 Quality: moderate 36% attrition 0% attrition in yoga, 6% attrition in exercise (attrition defined as attending no classes) 4% attrition Controls were crossed over into yoga 15% attrition in yoga, 20% attrition in control *P < 0.05; ** P < Pain effect size = Cohen s d comparing change scores of treated and control groups (or treated group only in uncontrolled studies). Quality = study quality designated high, moderate, low, or very low. AIMS 2 = Arthritis Impact Measurement Scales-2; BF = biofeedback; CES-D = Center for Epidemiologic Studies Depression scale; CPAQ = Chronic Pain Acceptance Questionnaire; EMG = Electromyographic; HPLP = Health Promoting Lifestyle Profile; LSI = life Satisfaction Index; K-WOMAC = Korean version of Western Ontario and McMaster University; MBSR = Mindfulness-Based Stress Reduction; MSCL = Medical Symptoms Checklist; MPQ = McGill Pain Questionnaire; MPQ-SF = McGill Pain Questionnaire short form; OA = osteoarthritis; RA = rheumatoid arthritis; SF-36 = Short Form 36-item Health Survey; SCL-90-R = Symptom Checklist 90-Revised; GSI = General Severity Index; VAS = visual analog scale; BAI = Beck Anxiety Inventory; CT = Cognitive Therapy; GDS = Geriatric Depression Scale; RCT = Randomized Controlled Trial; RDS = Roland Disability Scale; STAI = State-Trait Anxiety Inventory; Tx = Treatment.

7 Mind Body Pain Older Adult 365 PMR and then use a recording of the exercise for home practice. Over time, fewer and fewer muscle groups are tensed and released and, eventually, the person can relax the body at will without tensing beforehand. A combination of PMR, diaphragmatic breathing training, and biofeedback is often used in behavioral interventions for chronic pain in adults and children [17,20 23]. Few studies have evaluated PMR for chronic pain in an older population. In an uncontrolled pre post study, Arena et al. [24] reported on the effects of PMR therapy in 10 older tension headache patients aged years. Seventy percent of subjects showed >50% reduction in headache activity 3 months following the seven-session treatment, with two of the subjects reporting complete amelioration of headaches. The authors noted that subjects were asked to repeat back homework assignments at each session in order to facilitate understanding of the instructions. Lundgren and Stenstrom [25] assigned 68 rheumatoid arthritis patients (mean age 57) to 10 weeks of PMR training plus imagery (taught via audiotapes rather than personal instruction) or usual care. Although muscle function was found to improve, there was no decrease in pain or in disease activity. This study included follow-up assessments up to 1 year, at which time no significant difference between groups was found. In this study, the age range was broad, ranging from 28 to 70 years, and no information was provided regarding efficacy for older vs younger subjects. The authors noted that, although attendance was good and dropout rates were low, few participants practiced their relaxation skills at home as they had been instructed to do. Rybarczyk et al. [26,27] report on two randomized controlled studies of a multicomponent mind body wellness program for older adults (age 60 years) with chronic illness, such as osteoarthritis/rheumatoid arthritis (27% of the sample), spinal stenosis/low back pain (11%), and/ or diabetes (30%). The intervention included relaxation training (breathing, PMR, autogenic training, and imagery), information on mindfulness meditation, exercise, nutrition, and cognitive approaches. In the initial study, 178 subjects were randomly assigned to mind body wellness classes, a video-based home wellness program, or usual care control. Both treatment groups reported significantly greater pain reduction than the wait-list control group on the McGill Pain Questionnaire [28], as well as improved sleep. These authors subsequently compared 113 wellness class participants and 130 wait-list controls post treatment and at 1- year follow-up. At the 1-year follow-up, the sleep improvement was maintained, but pain improvement was not, perhaps indicating a need for maintenance contacts. Biofeedback The term biofeedback refers to a mind body therapy in which instrumentation is utilized to provide feedback to the patient regarding a physiologic process. Biofeedback has been used to help persons learn to regulate biological processes, such as heart rate, muscle tension, blood pressure, and vasoconstriction. Over time, the patient learns to control the biofeedback signal, which is typically an auditory or visual analog of the physiologic changes. The most frequent use of biofeedback is as a method of relaxation training, with the most frequently used methods being electromyographic (EMG) or muscle tension biofeedback, and thermal (peripheral skin temperature) biofeedback. Studies that evaluate the benefits of biofeedback for older persons with pain typically compare older vs younger participants. Frequently, biofeedback is one of several treatment components offered, along with other relaxation methods and cognitive therapy. This multimodal approach is typical in the clinic setting and most likely maximizes patient motivation and potential benefits. Therefore, differentiating the beneficial effects of biofeedback vs other forms of relaxation training is not possible. A wait-list controlled trial of biofeedback and relaxation for headache in older adults [29] was specifically tailored for older individuals, with a treatment length of 12 sessions. In this study, 14 participants aged years with either tension headache, migraine, or mixed headaches received EMG biofeedback training along with PMR and cognitive therapy. Subjects with migraines or mixed headache also received thermal biofeedback. Seven of these participants received the treatment after having completed a wait-list, diary-only condition. Although the small sample size precludes inferential statistical testing of relative benefits, the authors report that 2/7 of the control group and 7/14 of the biofeedback group exhibited improvement, which was defined as 50% reduction in headache activity. In this study, the treatment was adapted by lengthening sessions to 90 minutes and avoiding full muscle contraction during PMR in order to prevent discomfort. Several prospective but uncontrolled studies of biofeedback for headaches in the older adult sup-

8 366 Morone and Greco port the notion that older individuals can learn self-regulation skills to their benefit. Arena et al. [30] reported on a series of eight adults aged years with tension headache who received 12 sessions of biofeedback. Treatment sessions were modified slightly in order to facilitate comprehension; for example, sessions were slightly longer than is typical, and extra care was taken to evaluate whether subjects understood the concepts. Three months following the treatment, 50% of subjects had maintained significant decreases in headache activity, and three of the remaining four subjects reported moderate reductions in headache of 35 45%. In a prospective study of unstandardized behavioral treatment that included relaxation and biofeedback as well as cognitive strategies, 64% of the 16 older adults in the headache sample had greater than 50% improvement in headache at 1- month follow-up [31]. Multimodal, multidisciplinary chronic pain rehabilitation programs frequently include biofeedback training for relaxation, in addition to other forms of relaxation such as diaphragmatic breathing and PMR. Cognitive therapy and physical therapy are frequently offered in multidisciplinary programs as well [32,33]. In a series of publications based upon older and younger participants in a multidisciplinary chronic pain program, Middaugh and colleagues [34] found that older patients (55 78 years) did not have more difficulty than younger patients with the biofeedback/relaxation component of the overall program, and were consistent with younger pain patients in terms of finger temperature response to PMR training. Both older and younger persons were able to improve the ability to relax muscles at the site of pain (e.g., upper cervical area) in response to EMG biofeedback training, with 100% of the 20 younger patients and 75% of the 17 older patients meeting EMG success criteria. The older group reported significant reduction in pain from admission to discharge. In a further study with a larger sample size, Middaugh and Pawlick [35] compared 58 older (55 82 years) and 59 younger (18 54 years) participants in a multimodal pain treatment program who received biofeedback as well as CBT, medication management, and physical therapy. Although treatment components were consistent, some patients received inpatient treatment and some were outpatients. The number of biofeedback sessions ranged from 8 to 12. Patients received auditory or visual EMG biofeedback at the site of pain, such as the lumbar or trapezius muscles. Outcome measures included peripheral temperature change, respiration rate, and selfreport of pain. Older patients exhibited comparable results to younger persons, such as reductions in respiration rate and increase in skin temperature. The authors note, however, that older persons did not achieve the same end points in terms of temperature and respiration, but concluded that physical factors related to aging could account for this. Older patients reported greater reductions in pain ratings than younger patients. Although the design of the study precludes distinguishing the precise efficacy of biofeedback from other program components, the results support the idea that older adults with long-standing pain problems respond as well as younger individuals do to a multimodal pain treatment program that includes biofeedback. In summary, several biofeedback studies have addressed the ability of older adults with chronic pain to learn physiologic management, as well as benefit from this treatment. These reports suggest that the skills of physiologic management can be learned adequately by older chronic pain patients. Normal effects of aging on organs, such as decreased lung volume, may affect comparisons of outcome of older vs younger patients. Most of the studies reviewed are limited by small sample sizes, included multiple treatment modalities or did not include a control group. No randomized controlled trials that only study biofeedback within an older population have been reported. Concentration-Based Mind Body Therapies: Meditation, Guided Imagery, and Hypnosis Meditation Two common forms of meditation are mindfulness meditation and TM. Mindfulness meditation is taught in the mindfulness-based stress reduction (MBSR) program developed at the University of Massachusetts Medical Center and is now offered in many academic centers across the country. The program has also been operationalized and studied in many clinical trials. Simply, it involves bringing nonjudgmental moment-to-moment awareness to thoughts, sensations, or emotions as they arise. TM involves bringing awareness to a word or syllables that are continuously and silently repeated. While mindfulness meditation comes in different forms sitting meditation, walking meditation, loving-kindness meditation, or the body scan, TM typically involves one format. This is resting comfortably in a sitting position with the eyes closed while silently repeating the assigned word or syl-

9 Mind Body Pain Older Adult 367 lables. While the MBSR program in clinical trials typically involves weekly classes in a group format, TM in clinical trials also involves one-on-one initial instruction and some group classes. Both types of meditation require daily meditation practice, and both will provide guided meditation tapes to aid development of the meditation technique. As we have discovered while conducting this review, few studies have looked at older adults, and even fewer at older adults with persistent pain. Only two clinical trials of mindfulness meditation for persistent pain were found that included an older population, but none was found for TM. Thus, even though TM has been extensively studied in clinical trials [36,37] and has been shown to be feasible in an older population [38,39], the authors could not find any studies of TM for persistent pain in an older age group. Two recent reviews by Baer [40] and Grossman et al. [41] present compelling evidence for the benefit of the MBSR program in stress, anxiety, depression, and pain. An early study of Kabat- Zinn [42] examined the effects of the MBSR program on 58 subjects with chronic pain of varying etiology and with an age range of years. There was no control group. Subjects had significant improvement immediately after completing the program on measures of pain and mood. Morone et al. [43] conducted a preliminary study of 37 older adults (mean age of 74 years) with chronic low back pain and randomized them to an 8-week MBSR program or a wait-list control. The intervention was feasible among this population, with significant improvement in self-reported pain acceptance and physical function as compared with the control group. As with many of the mind body studies we review in this article, because of the small size of this pilot study, the results need to be interpreted with caution. Articles by McBee et al. [44] and Smith [45] describe the authors experience teaching MBSR to older adults and frail older adults. They both were able to successfully teach the program, but McBee et al. found that they needed to reduce the session time to 1 hour when working with the frail older adult in a nursing home setting. Guided Imagery Guided imagery sessions can occur in the presence of an individual trained in guided imagery, or can occur with audio tapes or CDs. Typically, a guided imagery session will begin with a relaxation exercise to focus attention and relieve tension of the patient before the actual guided imagery occurs. The initial exercise may involve a breathing exercise or visualizing a safe place. The session will then move into more specific guided images. Van Kuiken, in her 2004 meta-analysis [46], found four types of imagery that are practiced. One involves pleasant imagery such as imagery of a peaceful location. A second involves physiologically focused imagery imagery of fighting disease like white blood cell s attacking cancer cells. A third involves mental rehearsal such as successfully performing a task like public speaking. Mental reframing involves imagery that reinterprets a past experience and its associated emotions. The fourth type is receptive imagery that involves scanning the body for diagnostic or reflective purposes. There have been very few studies of guided imagery in older adults. Baird et al. [47] randomized 28 women with osteoarthritis who were at least 65 years of age to either 12 weeks of guided imagery with PMR or a control group. Participants were required to do two 10- to 15-minute guided imagery sessions with PMR twice a day. Eighteen were randomized to the intervention group and 10 to the control group (usual care). One subject dropped out of the intervention group, and 88.8% of subjects reported doing guided imagery at least once a day during the study period. There were significant reductions in pain and significant increase in mobility at the end of the trial. Because this study combined two modalities, it is impossible to separate out the effects of imagery from PMR. However, this study demonstrates the feasibility of guided imagery as an intervention in older adults with chronic pain and suggests benefits in pain reduction and increased mobility. Lewandowski [48] randomized 44 adults with chronic pain, defined as daily pain that was a level 4 or higher on the Wong Baker FACES [49] for at least the previous 3 months to a guided imagery intervention or a wait-list control. Subjects received a 7-minute guided imagery tape that they were to do three times a day for 4 days. The median age of the population was 61 years. Only two participants dropped out of the intervention group. Most participants practiced the guided imagery three times a day, as revealed in their diaries. The guided imagery group showed a significant reduction (P < 0.05) in the visual analog scale (VAS) as compared with the control group. This study was small, and guided imagery was conducted for a very short period of time. Nevertheless, the results also indicate that guided imagery is feasible in older adults.

10 368 Morone and Greco Hypnosis Hypnosis is defined by the American Society of Clinical Hypnosis as a state of inner absorption, concentration and focused attention [50]. Hypnosis is typically induced by a trained therapist. A hypnosis session usually has several components. It includes induction of the hypnotic state by the hypnotherapist who then gives suggestions once an adequate hypnotic state is induced, and then guides the patient out of the hypnotic state [51]. Sessions are typically individual, but can be in a group. Sessions vary in number, but may be as few as one or as many as a dozen. Patients can also be taught self-hypnosis and be provided with tapes for home practice. Hypnosis has been used to treat a wide variety of medical conditions for over 200 years. A recent study by Patterson and Jensen [52] reviewed randomized, controlled clinical trials of hypnosis for acute and chronic pain. Acute pain conditions included medical procedures such as bone marrow aspiration, burn wound dressing changes, labor pain, and a variety of surgical procedures. Hypnosis resulted in significant reductions in selfreported measures of pain as compared with no treatment, usual care, or an attention control in about half of the studies reviewed, for three studies it was as good as, but no better than, the above comparison conditions, and one study had mixed results. For chronic pain, self-reported measures of pain also significantly improved for conditions such as fibromyalgia and headache. However, it was noted that, in chronic pain conditions, hypnosis was not found to be superior to relaxation training or autogenic training. A meta-analysis in 2000 [53] looked at both experimentally induced pain and clinical pain conditions, and concluded that hypnosis was an effective analgesic. Another recent study of hypnosis for chronic pain by Jensen and Patterson [54] reviewed 19 controlled trials of hypnosis for chronic pain for such conditions as headache, cancer-related pain, fibromyalgia, osteoarthritis, low back pain, temporomandibular pain disorder, disability-related pain, and mixed chronic pain problems. The authors concluded that hypnosis resulted in more pain reduction than no treatment. Additionally, pain reduction was maintained up to 12 months of follow-up in some studies. Many of the trials reviewed above include older adults over the age of 65 years. However, we identified only one trial of hypnosis for pain that was targeted to an older population. This was the trial by Gay et al. [55] comparing hypnosis with relaxation to a wait-list control for osteoarthritis pain. Thirty-six subjects were randomized to one of the three groups. Participants in the hypnosis group received eight individual 30-minute Erikson hypnosis sessions. Participants in the relaxation group received eight individual sessions of guided Jacobson relaxation, which involved progressively tensing and relaxing muscles of the body. The main outcome variable was pain intensity using the VAS. It showed that both the hypnosis and relaxation group significantly improved, compared with the control group at 8 weeks. The hypnosis and relaxation group were not significantly different from each other. Only the hypnosis group maintained improvement at 3 months compared with the control, and there was no significant difference between any group at 6 months. As a result of the many clinical and experimental trials of hypnosis demonstrating significant benefit over a control condition, hypnosis has been supported by the American Psychiatric Association [56] and the NIH [14]. Many of the trials included older adults, but did not exclusively target them. Nevertheless, this is compelling evidence that hypnosis is feasible in older adults and can significantly reduce pain. Movement-Based Mind Body Therapies: Tai Chi, Qi Gong, and Yoga Tai Chi and Qi Gong Tai chi and qi gong are practices traditionally used in China for thousands of years and currently becoming popular lifestyle health practices as forms of low-impact exercise and stress management training in the West. Originating in Chinese martial arts, both tai chi and qi gong involve slow, controlled motions and focused breathing, and are thought to enhance the body s energy or qi/chi. Tai chi chuan, typically called tai chi in the United States, is widely practiced by older adults in China and has become increasingly popular in Western countries. Tai chi forms or exercises typically include a standard series of gentle fluid movements, mental concentration, and controlled breathing. Hartman et al. [57] assigned 33 communitydwelling adults with lower extremity osteoarthritis and mean age of 68 years to tai chi classes (totaling 24 hours) or a control group that consisted of usual care plus group meetings and telephone contacts. Quality of life, functional outcomes, and perceived arthritis self-efficacy were evaluated. Health satisfaction and overall arthritis selfefficacy were significantly improved in the tai

11 Mind Body Pain Older Adult 369 chi group compared with control. The differences between groups on pain and functional mobility measures (such as single-leg balance and chair rises) were in the predicted direction and approached, but did not reach statistical significance (P < 0.1). Although this study was limited by small sample size, several design strengths are worth mentioning. This study included an attentioncontrol group rather than wait-list or usual care alone, which supports the idea that the active treatment benefits are not accounted for merely due to time and attention of the therapist. The Hartman et al. study included objectively measured functional outcomes, such as chair rise and 50-foot walk, rather than solely self-reported function. One study using tai chi for older adults aged 55 years and above with chronic pain was found. This study focused on 72 women with osteoarthritis randomly assigned to tai chi or a telephonecontact control condition [58]. Joint pain and stiffness, as well as physical function, were significantly improved with treatment. However, the results of this study are difficult to interpret due to the high dropout rate (43% attrition in the tai chi participants). There are a number of tai chi studies in healthy older adults or those with conditions other than pain. Several of these include large numbers of participants and excellent study designs. A selection of these trials is described here in order to illustrate the feasibility and safety of this intervention for older adults, as well as the potential for preserving function in this vulnerable group. A number of clinical trials have evaluated tai chi chuan as an intervention to improve health and physical function and reduce risk factors for falls in older adults. Benefits to participants in a sample of 94 inactive subjects with a mean age of 73 years, included improved perceived physical function and exercise self-efficacy in comparison with a wait-list control [59]. Young et al. [60] assigned 62 subjects aged 60 years and over with hypertension to a 12-week tai chi or aerobic exercise program. Both groups exhibited significant reduction in blood pressure, and there was no difference between groups. Wolf et al. [61] randomly assigned 200 community-dwelling adults aged 70 years to tai chi, computerized balance training, or group education. Tai chi was associated with 47.5% reduced risk of multiple falls, and fear of falling was reduced relative to the education control group. Wolf and colleagues [62] also studied the effects of tai chi vs wellness education on physical function, heart rate, and blood pressure in 311 frail older adults (defined as more than 70 years old, in supervised living, and having experienced a fall within the last year). Chair-rise time, body mass index, heart rate, and systolic blood pressure decreased in the tai chi group, but increased among wellness education participants from baseline to 1-year follow-up. There are a small number of studies of tai chi in healthy older adults without chronic pain that evaluated pain as an outcome. Ross et al. [63] focused on balance, mood, and pain in their uncontrolled pilot study of 17 community-dwelling older adults aged years who were not taking pain medications. A statistically significant improvement in pain as assessed via a VAS was found, as well as improvement in self-rated mood. Notably, the participants pain level was low at pretreatment, averaging 1.5 on a 10-cm VAS scale. Irwin et al. [64] used a randomized wait-list control design to evaluate whether tai chi could increase herpes zoster virus (shingles) cell-mediated immunity as well as health status. The 36 participants were aged 60 years or over, had previous exposure to the herpes zoster virus, and were excluded if they had a chronic pain condition. Cell-mediated immunity was significantly increased after 15 weeks of three times weekly tai chi classes (tai chi group) and not significantly changed in the control group. The physical function scale and role-physical scales of the SF-36 were improved among those who completed the tai chi intervention relative to controls. Not surprisingly, tai chi was not associated with significant reduction in the bodily pain scale of the SF-36, possibly due to floor effects. Qi gong, like tai chi, is a traditional Chinese exercise practice involving gentle movement and focused breathing. Within the Chinese tradition, it is explained as a method to unblock the energy of the body, or qi, and facilitate its healthy flow. No studies were located that evaluated qi gong for chronic pain in older adults, and only one study was found that evaluated pain as an outcome measure in persons with chronic illness. Two studies of qi gong that included older adults are described here, as they support feasibility and safety as well as potential for improved function. Cheung et al. [65] evaluated the effects of Guolin qi gong, which includes slow walking, arm movements, and twisting of the trunk, on blood pressure in participants with mild hypertension. The age range was years. The 91 participants were randomly assigned to 22 hours of qi

12 370 Morone and Greco gong classes or conventional exercise that was matched for intensity. Both groups were assigned homework practice. Both groups exhibited statistically significant improvement in SF-36 bodily pain, with no relative benefit for either group. Significant blood pressure and heart rate decreases were also found in both groups. There was a 21% dropout rate among qi gong participants and 11% dropout rate in the exercise group. A recent study by Stenlund et al. [66] of 12 sessions of qi gong and group discussion vs usual care for 109 cardiac rehabilitation patients found improvements in perceived physical activity level as well as objectively measured balance and coordination among the qi gong participants, but not the control group. Pain was not measured in this study. Attrition was 14% among qi gong participants and 11% in controls. The scarcity of qi gong trials in the English-language literature in contrast to the greater number of tai chi studies may reflect the differing popularity of these approaches at the present time in the West. These approaches, which each combine low-impact controlled exercise, breathing, and meditative awareness, deserve study in older adults with chronic pain. Yoga Yoga originated from Indian culture over 2000 years ago. The type of yoga known as Hatha yoga has been popularized in American culture. It usually involves holding the body in a sequence of postures or asanas for a certain period of time, breathing exercises and meditation. The postures in Hatha yoga make up the bulk of a yoga session, which may typically last 1 2 hours. Frequently, the postures are done sequentially with the aim of increasing flexibility and strength. The breathing and meditation exercises are intended to calm and focus the mind and to develop greater awareness [67]. The possible therapeutic benefits of yoga have been studied for a wide array of medical conditions. (See Kirkwood et al. [68] for a systematic review of yoga for anxiety, and Raub [69] for a review of yoga s effects on musculoskeletal and cardiopulmonary function.) The growing popularity of yoga has led researchers to investigate yoga in older adults. An early trial by Haber in 1983 [70] investigated a 10- week yoga program in 106 older adults residing in two senior centers as compared with a control group that participated in popular group activities available at the senior center. Dropout rate was 16% of black participants and 22% of white participants. White, but not black, participants significantly improved on measures of hypertension and psychological well-being. Oken et al. [71] published a randomized controlled trial of yoga in healthy seniors. In total, 135 older adults with a mean age of 72 years were randomized to a 6- month weekly yoga class, exercise class, or a wait-list control group. The intervention was completed by 86% of the subjects in the yoga arm and 81% of subjects in the exercise arm. No significant change was found in the primary outcome measure of attention among the three groups. However, significant changes were found in measures of quality of life and the physical function measures of one-leg stand and chair sit and reach in the yoga group, but not the exercise group or the wait-list control group. An uncontrolled trial of Iyengar yoga for gait in 23 older adults found significant improvement in measures of peak hip extension and stride length [72]. Eighty-three percent of participants completed the intervention. Other preliminary studies of yoga in older adults have been conducted to look at the effects of yoga on hyperkyphosis [73]. These studies establish that yoga is feasible in an older population. Fewer studies have looked at yoga for the treatment of persistent pain in older adults. Kolanski et al. [74] studied yoga for symptoms of knee osteoarthritis in 11 older adults with a mean age of 58 years (range 50 68). The intervention was 8 weekly 1- to 1.5-hour Iyengar yoga sessions. Seven participants completed the intervention. They significantly improved on measures of pain and physical function on the WOMAC (Western Ontario and McMaster Universities) Osteoarthritis Index [75]. Garfinkel et al. [76] conducted a randomized controlled trial of yoga for treatment of osteoarthritis of the hands vs a waitlist control in 26 adults aged years. The intervention was 8 weekly 1-hour yoga sessions. One subject dropped out of the control group. Significant improvement was noted in joint tenderness and hand pain during activity. Three controlled trials of yoga for chronic low back pain included both younger and older adults [77 79]. Sherman et al. [77] and Williams et al. [78] found that yoga significantly improved measures of function and pain which persisted on 26- week and 3-month follow-up, respectively. A controlled trial of yoga for carpal tunnel syndrome also included a wide age range of subjects [80]. Compared with control, subjects significantly improved on measures of pain and grip strength.